Design Search Autocomplete System

Design an autocomplete system that returns top suggestions for the current prefix and learns completed queries. Examples Example 1: Input: queries = ["roblox studio","roblox","roadblocks","roblox studio","roblox game","roblox"], timestamps = [5,1,7,3,4,2], prefixes = ["rob","robl","road","x"] Output: [["roblox","roblox studio","roblox game"],["roblox","roblox studio","roblox game"],["roadblocks"],[]] Explanation: roblox and roblox studio each appear twice, so the earlier first timestamp puts roblox first. roblox game appears once. Example 2: Input: queries = ["alpha","alpine","alpha","alpine"], timestamps = [10,3,2,8], prefixes = ["al"] Output: [["alpha","alpine"]] Explanation: Both queries appear twice. alpha has earliest timestamp 2, while alpine has earliest timestamp 3. Example 3: Input: queries = ["aba","abb","aba","abb"], timestamps = [1,1,5,7], prefixes = ["ab"] Output: [["aba","abb"]] Explanation: The count and earliest timestamp both tie, so lexicographic order breaks the tie. Constraints queries.length == timestamps.length 0 <= queries.length <= 10^5 1 <= prefixes.length <= 10^4 1 <= queries[i].length <= 100 0 <= prefixes[i].length <= 100 0 <= timestamps[i] <= 10^9 queries[i] contains printable ASCII characters prefixes[i] is empty or contains printable ASCII characters